This is a revised grant application requesting continued funding of a residential cohort of 18,244 men in Shanghai, China, assembled during 1986-89 when subjects were between the ages of 45 and 64 years. At recruitment, all cohort members provided detailed dietary and medical histories as well as blood and urine specimens. The cohort has been followed for the occurrence of cancer and death through routine ascertainment of new cases from the population-based Shanghai Cancer Registry and Shanghai Vital Statistics Units, and annual visits to all known surviving cohort members. To date, the cohort has accumulated 204,010 person-years of observation. Only 231 subjects have been lost to follow-up. There have been 1,463 incident cancer cases and 2,599 cohort members have died. The leading cancer sites are lung, liver, stomach, and colorectum. Stroke accounts for 23% of all deaths. Active follow-up of this cohort will continue for another 5 years. During the annual personal visit to each surviving cohort member, buccal cells will be collected to facilitate future etiologic studies involving DNA analysis. Although study investigators have been successful in harvesting DNA from serum samples collected at baseline, cost-based analysis revealed that it is more cost-effective for future gene-based studies to rely on this new source of DNA. It is estimated that buccal cells will be collected from about 11,000 cohort subjects. During the next 5 years, a series of nested case-control studies will be conducted to further elucidate the interplay of genetic and dietary factors in influencing cancer risk. Specifically, study investigators will assess the roles of dietary aflatoxin and antioxidants (selenium, retinol, carotenoids, tea polyphenols) in cancers of the lung and liver; the role of dietary isothiocyanates in cancers of the stomach, esophagus, and colorectum; and the role of tea polyphenols in cancers of the stomach and esophagus. In all instances of diet-cancer investigations, relevant genes with potential modifying effects on the respective diet-cancer associations will be included in the investigations.